Super-resolving multiresolution images with band-independant geometry of multispectral pixels
This addresses resolution enhancement for remote sensing applications, but appears incremental as it builds on existing super-resolution techniques for multispectral data.
The paper tackles the problem of enhancing resolution for multispectral and multi-resolution images, such as Sentinel-2 satellite data, by separating band-dependent reflectance from band-independent geometry to propagate sub-pixel details, resulting in a method that preserves reflectance while improving detail.
A new resolution enhancement method is presented for multispectral and multi-resolution images, such as these provided by the Sentinel-2 satellites. Starting from the highest resolution bands, band-dependent information (reflectance) is separated from information that is common to all bands (geometry of scene elements). This model is then applied to unmix low-resolution bands, preserving their reflectance, while propagating band-independent information to preserve the sub-pixel details. A reference implementation is provided, with an application example for super-resolving Sentinel-2 data.